Surgery causes changes in hemostatic function that are associated with postoperative arterial and venous thrombotic complications. These complications cause significant morbidity and mortality. Inasmuch as perioperative changes in coagulation, fibrinolysis and platelet function appear to be modulated by the type of intraoperative anesthesia and postoperative analgesia, the potential exists for decreasing the incidence of these complications. This investigation explores the relationship between perioperative changes in hemostatic function and postoperative arterial thrombotic complications, and seeks to define the optimal anesthetic/analgesic regimen. We propose two clinical studies to answer the specific aims. The first is a double-blind clinical trial of 240 patients undergoing abdominal aortic surgery. Subjects will be randomized to one of four groups to receive either regional supplemented general anesthesia or general anesthesia alone during surgery, and either intravenous narcotic or epidural local anesthetic patient controlled analgesia postoperatively. All aspects of perioperative management will be standardized by protocol to insure consistent clinical care. Data collection will include 96 hours of continuous holter monitoring, serial ECG's, physical examinations and measurement of cardiac isoenzymes to monitor clinically important thrombotic events. Fibrinogen, prothrombin fragment (F1.2), fibrinolysis, and platelet reactivity will be followed throughout the perioperative period to assess hemostasis. The second is a double-blind study to evaluate the modulating effects of a local anesthetic and narcotic on stress hormone-induced changes in hemostasis. Twenty normal volunteers will receive three infusions (24-hour) of an epinephrine-cortisol-glucagon-vasopressin-angiotensin-II cocktail, in combination with intravenous bupivacaine, fentanyl, or placebo. This infusion paradigm increases circulating measures of hemostatic function and will enable us to determine whether bupivacaine and fentanyl modulate hemostasis. Whole blood will be assayed at baseline, 2, 8 and 24 hours for complete blood count, fibrinogen, platelet reactivity and fibrinolysis. With completion of these two studies, we hope to 1) determine the relationship between perioperative hemostatic function and postoperative arterial thrombotic complications, 2) elucidate the modulating effects of anesthetics and analgesics on hemostatic function and the incidence of postoperative arterial thrombotic complications, and 3) define an optimal perioperative regimen.